Re: [ferret_users] curvilinear data and coordinates units

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hi Bruno,
Thank you for the report.  This can be a difficult issue, treating the 
branch cut in the longitude coordinates correctly.  I believe what 
you're seeing stems from this change between Ferret v6.2 and v6.3,

http://ferret.pmel.noaa.gov/Ferret/documentation/release-notes/v6-3-release-notes#curvi_contours

It's often the case when redefining a variable, we forget to give the 
new variable the right units.  Thank you for the reminder.

Ansley


On 11/6/2012 3:04 AM, Bruno Levier wrote:
> Hi Ferreters
>
> Be careful when using curvilinear data to have the coordinates units 
> properly defined, otherwise the resulting plots can be faulty (Ferret 
> 6.82).
>
> Here is an example. I want to plot a global variable centered on the 
> Pacific Ocean (curvilinear grid ORCA). I define a new longitude.
>
> !! Longitude definition (without units)
> Let nav_lon_bis = IF nav_lon LT 24 THEN nav_lon + 360 ELSE nav_lon
> Shade my_var, nav_lon_bis, nav_lat
>
> The result (see fig1.gif) is bad.
>
> !! Longitude definition (with units)
> Let /Units=lon nav_lon_bis = IF nav_lon LT 24 THEN nav_lon + 360 ELSE 
> nav_lon
> Shade my_var, nav_lon_bis, nav_lat
>
> The result (see fig2.gif) is good.
>
> This problem happens with Ferret 6.82, but not with Ferret 6.2
>
> Bruno Levier